Detection Of Grassland Fungi

The question of what role is played by particular species or groups in relation to ecosystem function is fundamental to microbial ecology. The technological and conceptual challenge required by any attempt to answer this has led to an obsession with methods. For unit-restricted taxa (see Chapter 1) such as many dung fungi, it would appear to be a relatively simple question, though current data are based almost exclusively on basidiocarp presence. However, most fungal activity in grasslands takes place in the soil, the physicochemical complexity and small scale heterogeneity of which make it difficult to map the location of hyphae (Feeney et al., 2006). For terricolous basidiomycetes in particular, this presents a challenge, since their distribution can be addressed at a range of spatial scales from soil crumb to field level (from a few micrometres to many metres). For most species (excepting fairy ring-forming fungi) such detailed spatial information is largely absent, and without this information it is difficult to elucidate what resources are being decomposed by particular species.

Standard dilution plating seldom recovers basidiomycete colonies, mainly because they are slow-growing but also because their hyphae are tightly associated with soil particles (Warcup, 1951b; Thorn et al., 1996). However, Warcup (1959) was able to isolate several taxa from pasture soil and roots by plating soil crumbs or micromanipulating individual hyphae. Among the diverse bas-idiomycetes isolated by these methods were several resupinate taxa, including Peniophora and Athelia spp. (Warcup and Talbot, 1962), which only rarely fruit (on the underside of soil clods or in worm tunnels; Eriksson, 1949). Direct counts of fungal hyphae by microscopy have been informative with regard to fungal standing crop, showing the increase in fungal biomass in a grassland chrono-sequence following arable cultivation (van der Wal et al., 2006). Quantification of basidiomycete mycelium, identifiable to some degree by the presence of clamp connections, has been achieved in woodland systems (Frankland, 1982; Robinson et al. , 2005) but not to our knowledge in grasslands. Current biochemical approaches (e.g. ergosterol, phospholipid fatty acids (PLFA), etc.), while informative about overall fungal activity/biomass, are hitherto unable to dissect out the basidiomycete component. We refer the reader to the excellent review on the merits of these approaches by Robinson et al. (2005). The activity of saprotrophic basidiomycetes has also been investigated by study of lignolytic enzymes from grassland soils (Gramss, 1997).

Although lacking in specific biomarkers or reliable isolation methods, the study of basidiomycetes is distinctly advantaged by the fact that many species form macroscopic fruit bodies. Indeed, with the exception of a limited number of well-studied species, inferences about the ecology of basidiomycetes are largely derived from the spatiotemporal distribution of these reproductive structures. However, fruiting patterns of grassland fungi present if anything a greater challenge than those of woodland taxa since environmental conditions in grasslands are generally less conducive to basidiocarp formation and persistence, especially the often low and fluctuating levels of atmospheric humidity. In drier grasslands especially, fruiting data are very sparse (e.g. North American mycologists seldom conduct grassland forays; Leon Shernoff, personal communication), but recent data from molecular studies suggest that many of the species present fruit only very rarely (Lynch and Thorn, 2006).

Most data of basidiocarp occurrence are collected informally and non-quantitatively in fungus forays and thus are not easily interpretable in any ecological context. Gilbert's (1875) study of basidiomycetes in response to various agricultural treatments at Park Grass Rothamsted is probably the first systematic survey of grassland fungi, finding that rings of M. oreades were most abundant on plots treated with lime superphosphate (either alone or in combination with sodium and magnesium sulphates) and mostly absent from plots treated with N (ammonium or manured) or K. A broadly similar pattern was found for Hygrocybe spp., which were present in greatest diversity on untreated plots. Wilkins and Patrick (1939, 1940) were the first to apply a more quantitative approach, recording basidiocarp numbers in fixed quadrats (ca. 700 m2) visited repeatedly over 2 years. When assessing basidiomycete diversity in different habitat types, they found ca. 20% of the 620 species encountered were present in grassland compared to ca. 60% in deciduous woodland but only 38 spp. exclusive to grassland (e.g. Hygrocybe, Lycoperdon and Panaeolus spp.) and fewer species being found on clay soils compared to chalk or sand. The most common species at the 20 grassland sites was H. virginea, present on all soil types at '80-100% constancy'. After 70 years of agricultural intensification, it would be interesting to examine whether these fruiting patterns have changed at these sites. Arnolds (1989) found that diversity of grassland fungi was much greater in fields where there had been no addition of synthetic fertilizer, a finding confirmed by more recent surveying of permanent quadrats at a range of replicated grassland field experiments (Griffith et al., 2002, 2004). This is consistent with a decrease in the ratio of fungal:bacterial biomass (based on PLFA profiles) following fertilization (Bardgett et al., 1999).

The vagaries of basidiocarp production have been noted many times and several studies have illustrated discrepancies between patterns of fruiting and mycelial abundance below ground (Horton and Bruns, 2001). Even basidiocarp surveys repeated over several years may provide an incomplete picture of below-ground diversity (see Chapter 5; Parker-Rhodes, 1951), although information can be gathered for large areas in a very time- and cost-efficient manner. The potential pitfalls of basidiocarp surveys of grasslands are lucidly described by Arnolds (1992b) and Watling (1995), including consideration of differential longevity of basidiocarps, fruiting periodicity, annual fluctuations and succession.

DNA-based approaches have transformed our understanding of microbial ecology, for instance, with regard to ectomycorrhizal fungi in woodlands (Horton and Bruns, 2001; Lindahl et al., 2007; Chapter 10). The most useful data currently available are from sequencing of clone libraries based on PCR amplification with fungal-specific primers. These provide a useful snapshot of the species present, often revealing the presence of unexpected taxa (compared to basidiocarp data). Use of taxon-specific primers has revealed that basidiomycetes are two- to threefold less abundant (relative to total fungal abundance) in prairie grassland soil than woodland (Fierer et al., 2005; O'Brien et al., 2005). The most detailed study to date (Lynch and Thorn, 2006) identified almost 300 basidiomycete species in adjacent pasture and arable plots, with up to 9 species in some 10 g soil samples. These comprised 45 species of clavarioid fungi (20% of the total), as well as other taxa (e.g. Hygrocybe and Entoloma spp.) typically observed in oligotrophic grassland in Europe. Thus, the diversity revealed by genetic analysis greatly exceeded both the limited range of basidiocarps found at the site (http://lter.kbs.msu.edu/) and the 51 morphospecies isolated on selective media (Thorn et al., 1996). A similar disparity between molecular data, culture-based approaches and basidio-carp surveys was also observed in Welsh grasslands (Hunt et al., 2004).

Cloning and sequencing is costly when scaled up and more rapid fingerprinting approaches, such as terminal restriction fragment length polymorphism (T-RFLP) or fungal automated ribosomal intergenic spacer analysis (FARISA), can robustly reveal treatment effects, for example, along grassland fertilization gradients (Brodie et al., 2003; Kennedy et al., 2006). More powerful still is a dual approach allowing peaks in T-RFLP profiles to be identified from sequence data. However, the possibility of bias (due to primer specificity or differential efficiency of DNA extraction) can skew data (Anderson et al., 2003; Avis et al., 2006). A potential problem with genetic approaches relates to effective sampling, given the often very heterogeneous distribution of grassland basidiomycetes (Figure 2). One hectare of grassland contains ca. 1,000 t of topsoil (crudely assuming 10 cm soil depth and bulk density of 1 g cm-3) and it is very difficult to devise an effective sampling strategy to ensure representative coverage (when DNA extraction methods are limited to 1-10 g soil) without a very large budget. Technological advances, possibly soil fungus microarray chips (Sessitsch et al., 2006) or metagenomics, will increase efficiency of genetic approaches but basidiocarp surveys will remain a valuable complement of grassland research.

0 0

Post a comment